You will see VAERS numbers used two completely opposite ways: as proof something dangerous is being covered up, and as proof nothing is wrong at all. Both uses usually skip the same step, actually explaining what VAERS is and is not built to tell you. This is educational, not medical advice -- I want to walk through the tool plainly, so you can read a VAERS-based claim yourself instead of trusting whoever quoted the number at you.
What VAERS actually is
The Vaccine Adverse Event Reporting System is a passive surveillance database, co-managed by the FDA and CDC. 'Passive' is the operative word: it does not go out and look for problems. It waits for someone, a patient, a parent, a nurse, a doctor, anyone, to file a report describing something that happened after a vaccination. Filing a report requires no proof that the vaccine caused the event. It requires only a claim that the event happened afterward. That is by design -- VAERS exists to catch weak signals early, so a wide net beats a narrow one.
Why a report is not a finding
This is the piece that gets lost in both directions. A VAERS report records that someone experienced something and reported it near in time to a vaccination. It does not verify the event happened exactly as described, rule out every other cause, or establish that the vaccine was responsible. Temporal association, one thing happening after another, is not the same as causation, and VAERS was never built to make that distinction on its own. That limitation is stated by the system itself, not invented by critics -- and it is the reason researchers treat a raw VAERS count as a lead worth investigating, not a conclusion to report.
The honest gap: underreporting
Here is where the system's own limitations deserve real airtime, because they are documented by the government's own funded research, not by outside critics. An AHRQ-funded analysis carried out by Harvard Pilgrim Health Care, led by Dr. Richard Lazarus, built an electronic-health-record-based system to compare automated adverse-event detection against what actually reaches VAERS, and found that fewer than one percent of actual vaccine adverse events make it into VAERS through the existing passive reporting process.[1] That is a real, substantial, acknowledged gap in the surveillance system, published by the people who fund and run it. Being straight about that gap is not an anti-vaccine talking point. It is just accurate.
How a real signal actually gets evaluated
So what happens when a pattern in the data looks like it might be real? It does not stop at VAERS. The Institute of Medicine, now part of the National Academies of Sciences, Engineering, and Medicine, has run this exact evaluation multiple times over decades, and the process is public. In its most recent full review, an independent expert committee examined 158 specific vaccine-injury pairings people had raised and found convincing evidence supporting a causal link for 18 of them.[2] For the MMR-autism question specifically, working from the studies judged methodologically adequate (5 of the 22 identified), the committee's evidence favored rejecting a causal relationship. For some other pairings, the committee explicitly found the evidence inadequate to accept or reject causation either way, a real, honest category, not a dodge. An earlier version of the same review process, in 1994, found causal evidence for 12 of 54 claimed injuries and found the evidence inadequate to accept or reject causation for 38 more.[3] That range of outcomes, some confirmed, some rejected, some left open, is what an honest evaluation process looks like. A process that always concludes 'nothing to see here' would not be more trustworthy than one that sometimes finds something. It would be the opposite.
What real VAERS research looks like
It helps to see an actual example instead of an abstraction. Researchers published a peer-reviewed analysis comparing VAERS reports following an inactivated influenza vaccine in children under two, checking whether the thimerosal-free version of the vaccine produced a different pattern of reported reactions than the thimerosal-containing version. They found no difference in the proportion of injection-site reactions, rash, or reported infections between the two formulations.[4] Notice what that study is: a specific, checkable question asked of the VAERS data, with a specific, checkable answer, in this case a reassuring one. That is what proper use of VAERS looks like on either side of a debate: a defined question, not a raw number waved around as proof of a conclusion someone had already reached.
How to look it up yourself
You do not need anyone's summary, including this one. VAERS is public and searchable at vaers.hhs.gov -- you can search by vaccine, by age group, by reported event, and read the raw reports yourself. Every FDA-licensed vaccine has a package insert published at fda.gov, and the Adverse Reactions section lists what was reported in the trials and after release. Read the actual numbers next to each other before deciding what they mean. That habit is the whole point of informed consent: not trusting the headline in either direction, but going and looking.
Key Takeaways
- VAERS is a passive surveillance system -- anyone can file a report, and filing requires no proof the vaccine caused the event.
- A report is a claim of temporal association, not a verified causal finding -- that limitation is stated by the system itself, not invented by critics.
- A government-funded Harvard Pilgrim analysis (Lazarus et al.) found fewer than 1% of actual vaccine adverse events reach VAERS -- a real, acknowledged gap.
- When a signal needs real evaluation, an independent process exists: National Academies reviews have found causal evidence for some claimed injuries, rejected others, and honestly left some unresolved.
- VAERS is public and searchable at vaers.hhs.gov -- read the raw reports and the package insert yourself rather than trusting a number someone else quoted.
Sources
Powered by CellWell.life- 1.Lazarus R et al., Electronic Support for Public Health -- Vaccine Adverse Event Reporting System (ESP:VAERS), AHRQ Grant Final Report (2011)
- 2.Institute of Medicine, Adverse Effects of Vaccines: Evidence and Causality (2012)
- 3.Institute of Medicine, Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality (1994)
- 4.McMahon AW et al., Inactivated influenza vaccine (IIV) in children <2 years of age: examination of selected adverse events reported to VAERS after thimerosal-free or thimerosal-containing vaccine (2008)
Related
Want the next one of these?
The Signal is one email when there is something genuinely worth your attention -- written the way this post is, in plain language, with the studies behind it. No noise, no selling your address, and nothing to buy.
Spam protection is not configured for this form yet. Please use the email link instead.
The Signal, and nothing else -- no fixed schedule, only when there is something worth sending. We never sell or share your address, and you can unsubscribe in one click any time.